The impact of (simulated) resolution on breast cancer diagnosis based on high-resolution 3D micro-CT microcalcification images

BACKGROUND

Breast microcalcifications (MCs) are considered to be a robust marker of breast cancer. A machine learning model can provide breast cancer diagnosis based on properties of individual MCs - if their characteristics are captured at high resolution and in 3D.

PURPOSE

The main purpose of the study was to explore the impact of image resolution (8 µm, 16 µm, 32 µm, 64 µm) when diagnosing breast cancer using radiomics features extracted from individual high resolution 3D micro-CT MC images.

METHODS

Breast MCs extracted from 86 female patients were analyzed at four different spatial resolutions: 8 µm (original resolution) and 16 µm, 32 µm, 64 µm (simulated image resolutions). Radiomic features were extracted at each image resolution in an attempt, to find a compact feature signature allowing to distinguish benign and malignant MCs. Machine learning algorithms were used for classifying individual MCs and samples (i.e., patients). For sample diagnosis, a custom-based thresholding approach was used to combine individual MC results into sample results. We conducted classification experiments when using (a) the same MCs visible in 8 µm, 16 µm, 32 µm, and 64 µm resolution; (b) the same MCs visible in 8 µm, 16 µm, and 32 µm resolution; (c) the same MCs visible in 8 µm and 16 µm resolution; (d) all MCs visible in 8 µm, 16 µm, 32 µm, and 64 µm resolution. Accuracy, sensitivity, specificity, AUC, and F1 score were computed for each experiment.

RESULTS

The individual MC results yielded an accuracy of 77.27%, AUC of 83.83%, F1 score of 77.25%, sensitivity of 80.86%, and specificity of 72.2% at 8 µm resolution. For the individual MC classifications we report for the F1 scores: a 2.29% drop when using 16 µm instead of 8 µm, a 4.01% drop when using 32 µm instead of 8 µm, a 10.69% drop when using 64 µm instead of 8 µm. The sample results yielded an accuracy and F1 score of 81.4%, sensitivity of 80.43%, and specificity value of 82.5% at 8 µm. For the sample classifications we report for F1 score values: a 6.3% drop when using 16 µm instead of 8 µm, a 4.91% drop when using 32 µm instead of 8 µm, and a 6.3% drop when using 64 µm instead of 8 µm.

CONCLUSIONS

The highest classification results are obtained at the highest resolution (8 µm). If breast MCs characteristics could be visualized/captured in 3D at a higher resolution compared to what is used nowadays in digital mammograms (approximately 70 µm), breast cancer diagnosis will be improved.

The Upgrade Risk of B3 Lesions to (Pre)Invasive Breast Cancer After Diagnosis on Core Needle or Vacuum Assisted Biopsy. A Belgian National Cohort Study

INTRODUCTION

Flat epithelial atypia (FEA), lobular neoplasia (LN), papillary lesions (PL), radial scar (RS) and atypical ductal hyperplasia (ADH) are lesions of uncertain malignant potential and classified as B3 lesions by the European guidelines for quality assurance in breast cancer screening and diagnosis. Current management is usually wide local excision (WE), surveillance may be sufficient for some. We investigated the upgrade rate of B3 lesions to breast malignancy in a subsequent resection specimen after diagnosis on core needle-or vacuum assisted biopsy (CNB-VAB) in a national population-based series.

METHODS

Using data from the Belgian Cancer Registry (BCR) between January 1, 2013 and December 31, 2016, inclusion criteria were new diagnosis of a B3 lesion on CNB or VAB with subsequent histological assessment on a wider excision specimen. Histological agreement between first- and follow-up investigation was analyzed to determine the upgrade risk to ductal adenocarcinoma in situ (DCIS) or invasive breast cancer (IC) according to the type of B3 lesion.

RESULTS

Of 1855 diagnosed B3 lesions, 812 were included in this study: 551 after CNB-261 after VAB. After diagnosis on CNB and VAB, we found 19.0% and 14.9% upgrade to malignancy respectively. Upgrade risks after CNB and VAB were: FEA 39.5% and 17.6%; LN 40.5% and 4.3%; PL 10.4% and 12.5%; RS 25.7%and 0.0%; ADH 29.5% and 20.0%.

CONCLUSION

Based on the observed upgrade rate we propose three recommendations: first, resection of ADH, and FEA with WE; second, resection of RS and classical LN with therapeutic VAB and further surveillance when radio-pathological correlation is concordant; third, surveillance of PL.

Risk Severity Model for Pediatric Autosomal Dominant Polycystic Kidney Disease Using 3D Ultrasound Volumetry

BACKGROUND

Height-adjusted total kidney volume (htTKV) measured by imaging defined as Mayo Imaging Class (MIC) is a validated prognostic measure for autosomal dominant polycystic kidney disease (ADPKD) in adults to predict and stratify disease progression. However, no stratification tool is currently available in pediatric ADPKD. Because magnetic resonance imaging and computed tomography in children are difficult, we propose a novel 3D ultrasound-based pediatric Leuven Imaging Classification to complement the MIC.

METHODS

A prospective study cohort of 74 patients with genotyped ADPKD (37 female) was followed longitudinally with ultrasound, including 3D ultrasound, and they underwent in total 247 3D ultrasound assessments, with patients' median age (interquartile range [IQR]) at diagnosis of 3 (IQR, 0-9) years and at first 3D ultrasound evaluation of 10 (5-14) years. First, data matching was done to the published MIC classification, followed by subsequent optimization of parameters and model type.

RESULTS

PKD1 was confirmed in 70 patients (95%), PKD2 in three (4%), and glucosidase IIα unit only once (1%). Over these 247 evaluations, the median height was 143 (IQR, 122-166) cm and total kidney volume was 236 (IQR, 144-344) ml, leading to an htTKV of 161 (IQR, 117-208) ml/m. Applying the adult Mayo classification in children younger than 15 years strongly underestimated ADPKD severity, even with correction for height. We therefore optimized the model with our pediatric data and eventually validated it with data of young patients from Mayo Clinic and the Consortium for Radiologic Imaging Studies of Polycystic Kidney Disease used to establish the MIC.

CONCLUSIONS

We proposed a five-level Leuven Imaging Classification ADPKD pediatric model as a novel classification tool on the basis of patients' age and 3D ultrasound-htTKV for reliable discrimination of childhood ADPKD severity.

Retrospective Cohort Study of Practical Applications of Paramagnetic Seed Localisation in Breast Carcinoma and Other Malignancies

(1) Background: Paramagnetic seeds are a safe alternative for the wire-guided localisation of non-palpable breast lesions, but can also be applied for non-breast lesions. This study presents the experience with a paramagnetic seed, MagSeed® (Endomagnetics Ltd., Cambridge, UK, CE-registered and FDA-cleared), in an academic and non-academic breast centre. (2) Methods: Multicentre, retrospective analysis of 374 consecutive patients who underwent surgery after paramagnetic seed localisation (MSL) between 2018 and 2020. Indications for localisation included non-palpable breast lesions (n = 356), lymph nodes (n = 15) or soft tissue lesions (n = 3). The primary outcome was feasibility and the rate of positive section margins. The secondary outcome was predictive factors for positive section margins. (3) Results: The accurate excision of high-risk breast lesions, lymph nodes and soft tissue lesions was seen in 91.07% (n = 56). Positive section margins were observed in 7.86% (n = 25) after breast conserving surgery for invasive or ductal carcinoma in situ (DCIS) (n = 318). Invasive breast cancer associated with DCIS (p = 0.043) and the size of DCIS (p < 0.001) were significantly correlated with the positive section margins. (4) Conclusion: This study confirms the feasibility of MSL, as well as the higher risk for positive margins in cases of breast carcinoma with associated DCIS. Soft tissue lesions and lymph nodes associated with other malignancies, e.g., melanoma, can also be localised with paramagnetic seeds. This offers perspectives for future applications, such as the de-escalation of axillary treatment in breast cancer.

Towards updated understanding of brain metastasis

Brain metastasis (BM) is a common complication in cancer patients with advanced disease and attributes to treatment failure and final mortality. Currently there are several therapeutic options available; however these are only suitable for limited subpopulation: surgical resection or radiosurgery for cases with a limited number of lesions, targeted therapies for approximately 18% of patients, and immune checkpoint inhibitors with a response rate of 20-30%. Thus, there is a pressing need for development of novel diagnostic and therapeutic options. This overview article aims to provide research advances in disease model, targeted therapy, blood brain barrier (BBB) opening strategies, imaging and its incorporation with artificial intelligence, external radiotherapy, and internal targeted radionuclide theragnostics. Finally, a distinct type of BM, leptomeningeal metastasis is also covered.

Development and characterization of a rat brain metastatic tumor model by multiparametric magnetic resonance imaging and histomorphology

To facilitate the development of new brain metastasis (BM) treatment, an easy-to-use and clinically relevant animal model with imaging platform is needed. Rhabdomyosarcoma BM was induced in WAG/Rij rats. Post-implantation surveillance and characterizations were systematically performed with multiparametric MRI including 3D T1 and T2 weighted imaging, diffusion-weighted imaging (DWI), T1 and T2 mapping, and perfusion-weighted imaging (PWI), which were validated by postmortem digital radiography (DR), µCT angiography and histopathology. The translational potential was exemplified by the application of a vascular disrupting agent (VDA). BM was successfully induced in most rats of both genders (18/20). Multiparametric MRI revealed significantly higher T2 value, pre-contrast-enhanced (preCE) T1 value, DWI-derived apparent diffusion coefficient (ADC) and CE ratio, but a lower post-contrast-enhanced (postCE) T1 value in BM lesions than in adjacent brain (p < 0.01). PWI showed the dynamic and higher contrast agent uptake in the BM compared with the adjacent brain. DR, µCT and histopathology characterized the BM as hypervascular tumors. After VDA treatment, the BM showed drug-related perfusion changes and partial necrosis as evidenced by anatomical, functional MRI parameters and postmortem findings. The present BM model and imaging modalities represent a feasible and translational platform for developing BM-targeting therapeutics.

Development and characterization of a chick embryo chorioallantoic membrane (CAM) based platform for evaluation of vasoactive medications

Among various anti-cancer therapies, tumor vascular disrupting agents (VDAs) play a crucial role, for which their off-targeting effects on normal vessels need also to be investigated. The purpose of this study was to set up an in-ovo platform that combines a laser speckle contrast imaging (LSCI) modality with chick embryo chorioallantoic membrane (CAM) to real-time monitor vascular diameters and perfusion without and with intravascular injection. Two eggshell windows for both observation or measurement and injection were opened. Dynamic blood perfusion images and corresponding statistic graphs were acquired by using a LSCI unit on CAMs from embryo date (ED) 9 to ED15. A dedicated fine needle catheter was made for slow intravascular administration over 30 min with simultaneous LSCI acquisition. To verify the connectivity between CAM vessels and the embryonic circulations in the egg, contrast-enhanced 3D micro computed tomography (μCT), 2D angiography and histology were executed. This platform was successfully established to acquire, quantify and demonstrate vascular and hemodynamic information from the CAM. Chick embryos even with air cell opened remained alive from ED9 to ED15. Through collecting LSCI derived CAM vascular diameter and perfusion parameters, ED12 was determined as the best time window for vasoactive drug studies. A reverse correlation between CAM vessel diameter and blood perfusion rate was found (p < 0.002). Intravascular infusion and simultaneous LSCI acquisition for 30 min in ovo proved feasible. Contrast-enhanced angiography and histomorphology could characterize the connectivity between CAM vasculature and embryonic circulation. This LSCI-CAM platform was proved effective for investigating the in-ovo hemodynamics, which paves the road for further preclinical research on vasoactive medications including VDAs.

Prevalence, progression and implications of breast artery calcification in patients with chronic kidney disease

Breast arterial calcification (BAC) is increasingly recognized as a specific marker of medial calcification. The present retrospective observational cohort study aimed to define the prevalence, progression rate, risk factors and clinical implications of BAC in chronic kidney disease (CKD) patients across stages of disease. The presence and extent of BAC were determined on mammograms in 310 females (58.7 ± 10.8 years, Caucasian) with CKD across various stages of disease [CKD G2-5D n = 132; transplant (Tx) recipients n = 178]. In a subset of 88 patients, repeat mammography was performed, allowing us to calculate the annualized BAC rate. Overall, BAC was observed in 34.7% of the patients. BAC prevalence (P = 0.02) and BAC score (P = 0.05) increased along the progression of CKD. In the overall cohort, patients with BAC were characterized by older age, more cardiovascular disease, more inflammation, higher pulse pressure and borderline higher prevalence of diabetes and were more often treated with a vitamin K antagonist (VKA). The BAC progression rate was significantly lower in Tx patients as compared with CKD G5D. Progressors were characterized by more inflammation, worse kidney function, higher BAC score and higher serum phosphate level (Tx only) at baseline and were more often treated with a VKA. Major adverse cardiovascular event-free survival was significantly worse in Tx patients with BAC. In conclusion, BAC is common among CKD patients, progresses at a slower pace in Tx patients as compared with CKD 5D and associates with dismal cardiovascular outcomes. BAC score, kidney function, serum phosphate at baseline and VKA usage seem to be important determinants of progression.

How does image quality affect radiologists' perceived ability for image interpretation and lesion detection in digital mammography?

OBJECTIVES

To study how radiologists' perceived ability to interpret digital mammography (DM) images is affected by decreases in image quality.

METHODS

One view from 45 DM cases (including 30 cancers) was degraded to six levels each of two acquisition-related issues (lower spatial resolution and increased quantum noise) and three post-processing-related issues (lower and higher contrast and increased correlated noise) seen during clinical evaluation of DM systems. The images were shown to fifteen breast screening radiologists from five countries. Aware of lesion location, the radiologists selected the most-degraded mammogram (indexed from 1 (reference) to 7 (most degraded)) they still felt was acceptable for interpretation. The median selected index, per degradation type, was calculated separately for calcification and soft tissue (including normal) cases. Using the two-sided, non-parametric Mann-Whitney test, the median indices for each case and degradation type were compared.

RESULTS

Radiologists were not tolerant to increases (medians: 1.5 (calcifications) and 2 (soft tissue)) or decreases (median: 2, for both types) in contrast, but were more tolerant to correlated noise (median: 3, for both types). Increases in quantum noise were tolerated more for calcifications than for soft tissue cases (medians: 3 vs. 4, p = 0.02). Spatial resolution losses were considered less acceptable for calcification detection than for soft tissue cases (medians: 3.5 vs. 5, p = 0.001).

CONCLUSIONS

Perceived ability of radiologists for image interpretation in DM was affected not only by image acquisition-related issues but also by image post-processing issues, and some of those issues affected calcification cases more than soft tissue cases.

KEY POINTS

• Lower spatial resolution and increased quantum noise affected the radiologists' perceived ability to interpret calcification cases more than soft tissue lesion or normal cases. • Post-acquisition image processing-related effects, not only image acquisition-related effects, also impact the perceived ability of radiologists to interpret images and detect lesions. • In addition to current practices, post-acquisition image processing-related effects need to also be considered during the testing and evaluation of digital mammography systems.

Cancer surveillance in adults with germline TP53 pathogenic variants: A single-center observational study

10530

Background: Germline pathogenic variants (PV) in the tumor suppressor gene TP53 are associated with a high risk of developing diverse malignancies, often at young age, and predispose to Li-Fraumeni syndrome (LFS). Surveillance programs for presymptomatic PV carriers have shown survival benefit in a non-randomized trial. Here we describe the surveillance findings and clinical outcomes of adults with TP53 PV undergoing a standardized screening protocol. Methods: We identified adults with germline PV in TP53 who underwent surveillance at the University Hospitals Leuven, Belgium, between 04/2013 and 08/2020. Patients with prior cancer were allowed, while patients with an active malignancy requiring treatment at diagnosis of the TP53 PV were excluded. Surveillance was performed per modified Toronto protocol, including annual whole body diffusion-weighted MRI (WB-DWI/MRI), brain MRI, abdominal ultrasound (US), endoscopic surveillance, laboratory tests, dermatological examination and breast MRI/US in females. The primary aim was to evaluate the number and type of malignancies and premalignant lesions diagnosed during screening and to assess the proportion of malignancies detected by surveillance. Secondary outcomes were the cancer detection rate during the first year of screening, the proportion of carriers with false-positive findings, and overall survival. Results: We included 42 adults from 20 apparently unrelated families. Median age was 38y (range, 17-70y) and 23 had a history of prior cancer. After a median follow-up of 41.5mo, we diagnosed 18 cancers in 12/42 participants (29%). Overall survival was 95% in all participants, including 2 carriers who opted to discontinue surveillance. Surveillance detected 10/18 cancers (56%), the majority of whom through WB-DWI/MRI (6/10; 60%). No malignancies were identified with brain MRI. In 5/42 individuals (12%), surveillance detected a malignancy during the first year of screening. Only 2/10 cancers discovered with surveillance (1 soft tissue and 1 bone sarcoma) belong to the LFS core tumors. Cancers not detected with surveillance (8/18) were 6 non-melanoma skin cancers and 2 interval cancers (sarcoma post radiation, secondary acute leukemia). Additionally, we detected 27 premalignant lesions in 11/42 patients (26%), of whom 78% were diagnosed by colonoscopy. False-positive findings occurred in 7/42 patients (17%) and were mostly seen with WB-DWI/MRI. Conclusions: Adults with germline PV in TP53 that undergo surveillance have high cancer detection rates. The majority of malignancies were asymptomatic at diagnosis and detected with WB-DWI/MRI. Despite the high cancer incidence, few LFS core cancers were diagnosed and survival was encouraging. Increased genetic testing changes the clinical picture of germline TP53 carrier populations, justifying the transition from LFS to a wider concept of heritable TP53-related cancer syndrome.

Abstract PS7-32: Age of first full term pregnancy and other reproductive factors affect mammographic breast density in postmenopausal women

The objective of this study is to determine the role of age at first full term pregnancy (FFTP), and mammographic breast density (MBD) in postmenopausal women. For this purpose women, age 50-69y, participating in the Flemish (Belgium) population based breast cancer screening program were invited to participate in the study during their visit at the mammography department in the University Hospital Leuven (UZLeuven) or at the mobile unit of their hometown. A self-administered questionnaire providing detailed information on lifestyle and environmental exposure was completed by the participants. In this analysis 1,034 women were included (mean age of 59.1 years (range: 49.2 to 69.8). The three parameters of MBD were the percentage glandular tissue of the total breast volume (GLAND), the ratio of the glandular tissue volume compared to the whole breast volume or Volumetric Breast Density (VBD) and the BI-RADS density classification (VolparaDataManager®, USA). The first two parameters (GLAND and VBD) were transformed according to the natural logarithm to improve normality. These two parameters were linked with the FFTP via a piecewise linear regression (PLR) model using the NLIN procedure in SAS (version 9.4, SAS Institute, Cary, NC, USA). For the BI-RADS, we used ordinal logistic regression analysis. All models were corrected for a priori chosen variables: age at menarche, the ever use of oral contraceptive pill (OC), hormonal use at menopause, age at MBD measurement, and the body mass index at participation. Estimates are provided as a % change (95% confidence intervals [CI]).The average age at FFTP was 26.1 years (5th-95th percentile: 20 to 34). The PLR model estimated the breakpoint in our analysis at a FFTP of 25.7 years (95% CI: 22.3 to 29.1). For the woman with a FFTP younger than 25.7 years, the association between GLAND or VBD and FFTP is statistically not significant. For woman with a FFTP above 25.7 years, each year increase in FFTP was associated with 1.3% increase in GLAND (95% CI: 0.0% to 2.5%) and 1.5% increase in VBD (95% CI: 0.2% to 2.8%). Analysis of the BI-RADS showed similar results, the odds of belonging to a higher BI-RADS classification (e.g. from class 1 to class 2) increased with 5.4% (95% CI: 0.0% to 11.0%) for each year increase in FFTP age after the age of 25.7 years. Among other reproductive factors such as age at menarche, for every year delay there is 3.5% higher GLAND (95% CI: 1.4 % to 5.6%) and 2.5% higher VBD (95% CI: 0.5% to 4.6%) and the ever use of OC resulted in 10.0 % decrease in GLAND (95% CI: -18.8% to -0.2%) and 10.2% decrease in VBD (95%CI:-19.1% to -0.3%).In conclusion, this study is to our knowledge the first one to show that mammographic breast density is significantly increased when pregnancy takes place after 25.7 years of age. Later age at menarche is increasing and use of oral contraceptive is reducing mammographic breast density at postmenopause. (This study was supported by a Grant of the Breast Cancer Organization Think Pink of Belgium, the Department of Breast Radiology and Mobile Unit of the University Hospital Leuven, the Centre of Environmental Sciences and the Centre of Biostatistics of the Hasselt University, and NIH Grant CA06927 to FCCC, PA, USA).

Citation Format: Magda Johanna Vandeloo, Kristof Y. Neven, Liesbeth M. Bruckers, Jose Russo, Liesbeth Vancoillie, Esmee M. Bijnens, Eliane Kellen, Tim S. Nawrot, Chantal Van Ongeval. Age of first full term pregnancy and other reproductive factors affect mammographic breast density in postmenopausal women [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS7-32.

Abstract PS7-52: The upgrade risk to (pre-)invasive breast cancer for B3 lesions diagnosed on core needle or vacuum assisted biopsy. A Belgian retrospective study

Introduction: Flat epithelial atypia (FEA), classical lobular neoplasia (LN), papillary lesions (PL), radial scars (RS) and atypical ductal hyperplasia (ADH) are lesions of uncertain malignant potential in the breast, which are described as B3 lesions in the B classification system of the European Working Group for Breast Cancer Screening Pathology. Current standard for management of B3 lesions on core needle biopsy (CNB) or vacuum assisted biopsy (VAB), is wide local excision (WE). However, recent consensus-based guidelines no longer recommend WE for all such lesions, but propose surveillance following CNB or VAB to be sufficient in some cases. In the absence of a Belgian guideline on the treatment of B3 lesions, this study was conducted to identify which B3 lesions have the lowest likelihood for breast malignancy and could therefore be spared from WE.

Methods: Using data from the Belgian Cancer Registry (BCR), all patients with a new diagnosis of a B3 lesion on CNB or VAB between 2013-2016 and who had a histological follow up with VAB or WE after CNB or WE after VAB were included. Histological follow-up was retrieved from BCR and limited to 12 months following diagnosis. Histology was compared between the first- and follow-up investigation to determine the upgrade risk to ductal carcinoma in situ (DCIS) or invasive breast cancer (IC) according to the type of B3 lesion. Patients with synchronous (pre-) invasive lesions were excluded.

Results: Between 2013-2016 there were 812 B3 lesions available for upgrade analysis after initial diagnosis. After CNB 551 lesions had WE or VAB as follow up and after VAB 261 lesions had WE. After primary diagnosis on CNB, the total upgrade risk was 19,0%. There was histological agreement in 57,9% and no B3 lesion or upgrade was reported in 21,8%. Per B3 lesion subtype the upgrade risk to DCIS - IC after diagnosis on CNB was: ADH 17,1%-12,4%, FEA 21,1% - 18,4%, LN 18,9% -21,6%, RS 14,3% - 11,4%, and PL 7,2% - 3,2%. After initial diagnosis on VAB the total upgrade risk was 14,9%. There was histological agreement in 52,9% and no B3 lesion or upgrade was found in 31,4%. Per B3 subtype the upgrade risk to DCIS - IC after diagnosis on VAB was: ADH 17,3%- 2,7%, FEA 11,7%- 5,9%, LN 0,0% - 4,3%, PL 10,4% - 2,1%. We found no upgrade for RS. (Table 1).

Conclusions:

In a series of B3 lesions with a histological follow-up, we notice that overall upgrade risk is higher for lesions detected on CNB than on VAB: 19,0% vs. 14,9%. The majority of lesions showed histological agreement between initial B3 diagnosis and histological follow-up: 57,9% after CNB and 53,5% after VAB. More investigation is needed to make a proper risk assessment as to which B3 lesions can be followed with regular surveillance. Also, further prospective research is needed to get a better understanding of associated risk factors for upgrade, upgrade risk and lifetime risk of developing breast cancer after diagnosis of a B3 lesion.

Table 1: Upgrade risk per B3 subtype after diagnosis on CNB or VABSubtypeNumberCNB followed by VAB or WEDCISICTotal Upgrade riskCNBNumberVAB followed by WEDCISICUpgrade riskVAB followed by WEADH10517,1%12,4%29,5%11017,3%2,7%20,0%FEA3821,1%18,4%39,5%5111,7%5,9%17,6%LN3718,9%21,6%40,5%460%4,3%4,3%PL3367,2%3,2%10,4%4810,4%2,1%12,5%RS3514,3%11,4%25,7%60%0%0%Total551261Upgrade105/55119,0%39/26114,9%B3 lesion319/55157,9%138/26152,9%No B3 lesion120/55121,8%82/26131,4%Result not available7/5511,3%2/2610,8%

Citation Format: Nynke Willers, Patrick Neven, Giuseppe Floris, Cecile Colpaert, Eva Oldenburger, Sileny Han, Chantal Van Ongeval, Ann Smeets, Francois Duhoux, Hans Wildiers, Petra Denolf, Isabel De Brabander, Nancy Van Damme, Harlinde De Schutter. The upgrade risk to (pre-)invasive breast cancer for B3 lesions diagnosed on core needle or vacuum assisted biopsy. A Belgian retrospective study [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS7-52.

Development and content validity evaluation of a candidate instrument to assess image quality in digital mammography: A mixed-method study

PURPOSE

To develop a candidate instrument to assess image quality in digital mammography, by identifying clinically relevant features in images that are affected by lower image quality.

METHODS

Interviews with fifteen expert breast radiologists from five countries were conducted and analysed by using adapted directed content analysis. During these interviews, 45 mammographic cases, containing 44 lesions (30 cancers, 14 benign findings), and 5 normal cases, were shown with varying image quality. The interviews were performed to identify the structures from breast tissue and lesions relevant for image interpretation, and to investigate how image quality affected the visibility of those structures. The interview findings were used to develop tentative items, which were evaluated in terms of wording, understandability, and ambiguity with expert breast radiologists. The relevance of the tentative items was evaluated using the content validity index (CVI) and modified kappa index (k*).

RESULTS

Twelve content areas, representing the content of image quality in digital mammography, emerged from the interviews and were converted into 29 tentative items. Fourteen of these items demonstrated excellent CVI ≥ 0.78 (k* > 0.74), one showed good CVI < 0.78 (0.60 ≤ k* ≤ 0.74), while fourteen were of fair or poor CVI < 0.78 (k* ≤ 0.59). In total, nine items were deleted and five were revised or combined resulting in 18 items.

CONCLUSIONS

By following a mixed-method methodology, a candidate instrument was developed that may be used to characterise the clinically-relevant impact that image quality variations can have on digital mammography.

Cancer Surveillance in Healthy Carriers of Germline Pathogenic Variants in BRCA1/2: A Review of Secondary Prevention Guidelines

Germline pathogenic alterations in the breast cancer susceptibility genes 1 (BRCA1) and 2 (BRCA2) are the most prevalent causes of hereditary breast and ovarian cancer. The increasing trend in proportion of cancer patients undergoing genetic testing, followed by predictive testing in families of new index patients, results in a significant increase of healthy germline BRCA1/2 mutation carriers who are at increased risk for breast, ovarian, and other BRCA-related cancers. This review aims to give an overview of available screening guidelines for female and male carriers of pathogenic or likely pathogenic germline BRCA1/2 variants per cancer type, incorporating malignancies that are more or less recently well correlated with BRCA1/2. We selected guidelines from national/international organizations and/or professional associations that were published or updated between January 1, 2015, and February 1, 2020. In total, 12 guidelines were included. This review reveals several significant discordances between the different guidelines. Optimal surveillance strategies depend on accurate age-specific cancer risk estimates, which are not reliably available for all BRCA-related cancers. Up-to-date national or international consensus guidelines are of utmost importance to harmonize counseling and proposed surveillance strategies for BRCA1/2 carriers.

Abstract P6-08-03: Germline mutational landscape in 5422 individuals at risk for hereditary breast and ovarian cancer who underwent multi-gene panel testing

Background: The introduction of multi-gene panel testing and improved awareness under patients and physicians has led to an increase of individuals with known germline pathogenic variants in hereditary breast and ovarian cancer (HBOC) genes. Significant regional differences exist in germline mutational landscape. We aimed to report the findings from multi-gene panel testing in a large Belgian cohort of individuals at risk for HBOC.

Methods: All individuals who underwent multi-gene panel testing for HBOC at the Center for Human Genetics of the University Hospitals Leuven since the introduction of the panel were included (March 2016-April 2019). All included individuals were considered candidates for HBOC-panel testing by the requesting physician based on a personal or familial history of breast and/or ovarian cancer. Testing criteria from the Belgian Society of Human Genetics (www.beshg.be/download/guidelines/Guidelines_HBOC_2018.pdf) were met in the vast majority. The panel used was the BRCA Hereditary Cancer MASTR Plus® (Agilent, Belgium), with sequencing of BARD1, BRCA1, BRCA2, BRIP1, RAD51C, RAD51D, TP53, MRE11A, RAD50, NBN, FAM175A, ATM, PALB2, STK11, MEN1, PTEN, CDH1, MUTYH, CHEK2, BLM, XRCC2, EPCAM, MLH1, MSH6, PMS2 and MSH2. Sequencing was performed by NGS on a Miseq platform (Illumina). Genomic deletions and duplications in BRCA1 and BRCA2 were investigated with multiplex ligation-dependent probe amplification.

We hereby report on the frequency of pathogenic and likely pathogenic germline variants in this population.

Results: In 5422 individuals who underwent multi-gene panel testing, we detected 665 pathogenic or likely pathogenic variants in 639 patients (11,7%). In 25 patients (0.46%), more than one relevant alteration was detected with double heterozygosity in 24 individuals and triple heterozygosity in one. Germline variants in BRCA1 and BRCA2 were detected in 178 (3.3%) and 144 (2.7%) patients, resulting in a fraction of 26,4% and 21,4% of detected variants respectively. Relevant alterations in CHEK2, ATM, PALB2 and TP53 were observed in 135 (2.5%), 93 (1.7%), 26 (0.5%) and 11 (0.2%) patients respectively, accounting together for 39.3% of detected variants. Alterations in BRIP1/RAD51C/RAD51D were retrieved in 64 patients (1.2%) and alterations in mismatch-repair genes MSH6/MLH1/PMS2/MSH2 were detected in 0.3% of patients. These patients where dominantly referred for familial history of ovarian cancer. Furthermore, germline alterations in PTEN, CDH1 and BLM were observed in 3, 2 and 1 cases respectively. Double heterozygosity for ATM+CHEK2 and for ATM+BRCA2 were both observed in 3 cases. In the patient with triple heterozygosity, co-occurrence of pathogenic variants in BRCA2, ATM and CHEK2 was detected.

Conclusions: In a large Belgian cohort of 5422 individuals at risk for HBOC who underwent multi-gene panel testing, a pathogenic or likely pathogenic germline variant was detected in 11,7% of patients, and in 0,46% of patients double or triple heterozygosity for HBOC-variants was observed. Almost 40% of detected variants were alterations beyond BRCA correlated with hereditary breast cancer (CHEK2, ATM, PALB2 and TP53). Given the time-lag to predictive testing in families, a significant rise in healthy carriers with these non-BRCA alterations is expected in the upcoming years.

Citation Format: Kevin Punie, Griet Hoste, Griet Van Buggenhout, Ellen Denayer, Hilde Brems, Hilde Peeters, Ann Smeets, Ines Nevelsteen, Patrick Neven, Jan Ardui, Renate Prevos, Machteld Keupers, Chantal Van Ongeval, Giuseppe Floris, Christine Desmedt, Hans Wildiers, Geneviève Michils, Hilde Van Esch, Eric Legius. Germline mutational landscape in 5422 individuals at risk for hereditary breast and ovarian cancer who underwent multi-gene panel testing [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P6-08-03.

Abstract P3-08-31: Clinical and pathological features of invasive micropapillary carcinoma of the breast and correlation with prognosis

Background and objective: Invasive micropapillary carcinoma (IMPC) of the breast is a special variant of breast carcinoma with a unique morphology. Data about long-term outcome are conflicting and reports on stromal tumor infiltrating lymphocytes (sTILs) and their correlation with prognosis are scarce. In this retrospective cohort study we aimed to describe clinical and pathological features of IMPC, including sTILs assessment and immunohistochemistry studies, and their correlation with long-term outcome. Materials and methods: Patients with stage I-III IMPC (pure and mixed forms) who underwent upfront surgery at our institution between 2000 and 2016 were included. All patients signed informed consent. Standard clinico-pathological features and follow-up data to calculate distant relapse-free interval and breast cancer-specific survival were obtained from clinical records. Pathologic review of representative H&E-slides of the resection specimens included evaluation of sTILs and assessment of the micropapillary component. Surrogate molecular subtypes were based on receptor-status and histological grade. Using tissue microarrays we assessed by immunohistochemistry the pattern of staining of P53, and scored semi-qualitatively the expression of Bcl2, PAX8 and WT1. The association between predictors and outcome is analyzed using the Fine and Gray model, accounting for other-cause death as competing event. All tests are two-sided, assuming a 5% significance level. The sample size did not allow multivariate analysis. Results: We included 111 patients (median age 61,5 years; range 33-88). Luminal surrogate subtypes were most prevalent with 51 luminal A-like, 41 luminal B-like, 12 luminal HER2+, 5 HER2+ and 2 triple negative IMPC. 89% of patients had a T1 or T2 tumor and 50% of IMPC were poorly differentiated. Lymph node involvement was present in 59% and lymphovascular invasion in 44% of cases. Adjuvant chemotherapy, radiotherapy and endocrine therapy was administered in 47%, 87% and 90% of patients, respectively. Of all cases 59% were pure IMPC. Standard clinico-pathological features were comparable between pure and non-pure IMPC. sTILs were classified as low (&lt;30%), intermediate (30-50%) and high (&gt;50%) in 78%, 14% and 8% of specimens respectively. Comparison between surrogate subtypes showed higher sTILs (p=0.025) and a higher likelihood of aberrant P53 expression (p&lt;0.001) in HER2+ compared to luminal A-like subtype. Immunohistochemistry studies performed on 105 samples with enough material showed aberrant P53 expression in 10% and WT1 nuclear expression in 7% of cases. Pax8-staining was negative in all IMPC in this cohort. Bcl-2 expression was strongly related to all luminal subtypes (p&lt;0.03). After a median follow-up of 100 months, we observed 8 distant relapses (7,2%) and 3 breast cancer-related deaths (2,7%). All events occurred in non-pure IMPC. Surrogate subtypes for patients with distant relapses where luminal A-like in 4 patients while the other subtypes where each observed in 1 patient. Five out of eight patients with distant relapse had received prior adjuvant chemotherapy. Six had lymph node involvement. Higher median sTILs was correlated with worse distant relapse-free interval (HR=1.55; p=0.0172) and breast cancer-specific survival (HR=2.10; p&lt;0.001). Conclusions: Standard clinico-pathological features were similar in pure and non-pure IMPC. Despite high proportion of grade 3 differentiation and lymph node involvement, we observed a low rate of distant metastasis and within the pure IMPC (59% of patients), no distant relapses occurred. These findings could be explained by the high proportion of luminal-A like tumors in our cohort and need confirmation. Higher sTILs was associated with worse outcome in this IMPC cohort, confirming previously published observations.

Citation Format: Kevin Punie, Frederik Deman, Annouschka Laenen, Timothy Faes, Hans Wildiers, Ann Smeets, Ines Nevelsteen, Chantal Van Ongeval, Adinda Baten, Melissa Christiaens, Eva Oldenburger, Hilde Janssen, Caroline Weltens, Tatjana Geukens, Nynke Willers, Jan Ardui, Hava Izci, Laurence Slembrouck, Patrick Neven, Christine Desmedt, Giuseppe Floris. Clinical and pathological features of invasive micropapillary carcinoma of the breast and correlation with prognosis [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P3-08-31.

Abstract P1-10-07: The presence of ductal carcinoma in situ in core needle biopsy and microcalcifications on mammography in TNBC is associated with a lower pCR and worse long term outcome

Purpose Predictors for pathological Complete Response (pCR) in TNBC are very important given pCR is considered a surrogate marker for breast cancer-related survival. Associated DCIS in TNBC (accom-DCIS) as well as presence of microcalcifications on initial mammography (micro-MG) have been correlated with Androgen Receptor positive TNBC, known to achieve lower pCR rates. We aim to investigate the predictive impact of accom-DCIS and micro-MG on pCR. We also validated known clinico-pathological predictors for pCR. Long term outcome was analyzed for pCR and accom-DCIS. Methods A retrospective cohort study (diagnosed 1/1/2000 - 31/12/2017) of prospectively registered consecutively treated TNBC patients was performed. TNBC was defined as IHC-ER&lt;1%, IHC-PR&lt;1% and IHC-HER2 0-1+ and FISH negative if HER2 IHC 2/3+. All patients had standard of care neoadjuvant chemotherapy (NAC) in our hospital. Patients with at least 2 years of documented follow up and only TNBC of no special type on core needle biopsy (CNB) were included. We evaluated the predictive value of patient factors (age, BMI, symptomatic/screening-detected), imaging (micro-MG) and tumor related factors as reported in the pathology-report (grade, cT, focality, cN, lymphovascular invasion, accom-DCIS) for pCR (defined as pT0-TisN0). We used distant relapse rate and death to evaluate outcome. The Cox proportional hazards model was used to analyze the effect of pCR and accom-DCIS on distant relapse rate as well as overall survival (OS) in a multivariate model (corrected for size, nodal state, grading). Results are presented as hazard ratios (HR) with 95% confidence intervals (CI). Results The study contains 219 patients; accom-DCIS was reported in 53/219 (24.2%) CNB’s; pCR was achieved in 90/219 (41,1%). Patient related factors were comparable in patients who achieved pCR and those who had residual disease. Unifocal, smaller clinical size tumors and high tumor grade were significant predictors for pCR in our series. Analysis of 194 available mammographies showed micro-MG in 63/194 (32.5%). Micro-MG predicted for residual disease after NAC [HR = 3.350; 95% CI (1.708 - 6.569), p=0.005]. Accom-DCIS was a strong negative predictor of pCR [HR = 3.333; 95% CI (1.558 - 7.143), p=0.002)] and was associated with more distant relapse and worse survival [HR = 2.664; 95% CI (1.575 - 4.505), p=0.003]. pCR was strongly associated with a lower distant relapse rate and a better OS [HR = 2.210; 95% CI (1.282 - n3.811), p=0.004]. In the 129 cases without pCR, remaining tumor size [HR = 1.019; 95% CI (1.012 - 1.025), p=&lt; 0,001), ypN [HR = 1.917; 95% CI (1.539 - 2.388), p=&lt; 0,001) and presence of LVI [HR = 3.720; 95% CI (2.057 - 6.728), p=&lt; 0,001] significantly predicted for secondary metastasis. There was only a trend towards more distant relapses if accom-DCIS was found in the resection specimen in those who had residual disease [HR = 1.347; 95% CI (0.780 - 2.325), p=0.2855]. Conclusion Presence of accom-DCIS in CNB as well as micro-MG predicts for less pCR in TNBC. Accom-DCIS in CNB is associated with more distant relapse and worse OS. For accom-DCIS on resection specimen, there was only a tendency towards more distant relapses.

Citation Format: Jan Ardui, Sophie Vandamme, Chantal Van Ongeval, Giuseppe Floris, Hava Izci, Hans Wildiers, Kevin Punie, Tatjana Geukens, Ignace Vergote, Patrick Berteloot, Toon Van Gorp, Ann Smeets, Els Van Nieuwenhuysen, Sileny Han, Annouschka Laenen, Caroline Weltens, Hilde Janssens, Patrick Neven. The presence of ductal carcinoma in situ in core needle biopsy and microcalcifications on mammography in TNBC is associated with a lower pCR and worse long term outcome [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P1-10-07.

Abstract P1-20-24: Magnetic seed localisation as the new standard of care for nonpalpable breast lesion localisation: A comparison with hooked-wire localisation

Background and aims: Breast cancer is the most frequent type of cancer diagnosed in women. In breast surgery, screening programs and high-resolution imaging result in higher incidence of impalpable breast lesions, requiring preoperative localisation. Several localisation methods have been developed during the past decades. Since September 2018, Magseed® localisation has become standard of care in University Hospitals Leuven, providing an alternative for the widely used hooked-wire localisation method. The aim of this observational study was to compare hooked-wire localisation and Magseed® localisation in terms of oncological safety, clinical safety and surgeon satisfaction. Methods: Data of 100 patients who underwent Magseed® localisation were prospectively collected between September 2018 and April 2019, and retrospectively of 91 patients who underwent hooked-wire localisation between March 2018 and September 2018. Three patients received two Magseeds® resulting in a total of 103 seeds. Similarly, 11 patients received two hooked-wires so 102 wires were included in this study. Results: Baseline patient and tumour characteristics were similar between both groups. All magnetic seeds were placed under ultrasound guidance, with a median of two days in advance, resulting in logistical ease and more patient comfort. Occurrence of haematomas did not significantly differ between Magseed® and hooked-wire placement (2.97% vs 2.13% ; p = 1.000). 96.12% of the magnetic seeds were placed within 5 mm of the target, similar to 98.04% of the hooked-wires (p = 0.683). All seeds were retrieved during surgery, with an unambiguous detection by Sentimag® in 94.06%. The median ratio excised specimen volume to initial breast volume differs significantly between both groups in favour of hooked-wire (4.5% vs 3.7% ; p = 0.049), but the specimen volume itself is almost identical between the two groups. Positive surgical margin rate was lower for Magseed® compared to hooked-wire, although not significant (4.76% vs 10.39% ; p = 0.233). Due to positive margins, one additional mastectomy was performed in the Magseed® group. In contrast, three mastectomies and one additional excision were necessary in the hooked-wire group. 81% of the Magseed® procedures were scored as ‘easier than hooked-wire’ by our surgeons. Conclusion: The positive margin rate and associated re-excision/mastectomy rate was lower with Magseed® localisation compared to hooked-wire localisation. In addition, comparison of the complication rate showed an equal clinical safety profile. Also specimen volumes are similar with both localisation techniques. Moreover, the high surgeon satisfaction and logistical advantages designate Magseed® localisation as preferable over hooked-wire localisation.

Citation Format: Amelia Dehaene, Ann Smeets, Hanne Vos, Chantal Van Ongeval, Patrick Neven, Annouschka Laenen, Renate Prevos, Thomas Thywissen, Machteld Keupers, Ines Nevelsteen. Magnetic seed localisation as the new standard of care for nonpalpable breast lesion localisation: A comparison with hooked-wire localisation [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P1-20-24.

Image-guided breast biopsy and localisation: recommendations for information to women and referring physicians by the European Society of Breast Imaging

We summarise here the information to be provided to women and referring physicians about percutaneous breast biopsy and lesion localisation under imaging guidance. After explaining why a preoperative diagnosis with a percutaneous biopsy is preferred to surgical biopsy, we illustrate the criteria used by radiologists for choosing the most appropriate combination of device type for sampling and imaging technique for guidance. Then, we describe the commonly used devices, from fine-needle sampling to tissue biopsy with larger needles, namely core needle biopsy and vacuum-assisted biopsy, and how mammography, digital breast tomosynthesis, ultrasound, or magnetic resonance imaging work for targeting the lesion for sampling or localisation. The differences among the techniques available for localisation (carbon marking, metallic wire, radiotracer injection, radioactive seed, and magnetic seed localisation) are illustrated. Type and rate of possible complications are described and the issue of concomitant antiplatelet or anticoagulant therapy is also addressed. The importance of pathological-radiological correlation is highlighted: when evaluating the results of any needle sampling, the radiologist must check the concordance between the cytology/pathology report of the sample and the radiological appearance of the biopsied lesion. We recommend that special attention is paid to a proper and tactful approach when communicating to the woman the need for tissue sampling as well as the possibility of cancer diagnosis, repeat tissue sampling, and or even surgery when tissue sampling shows a lesion with uncertain malignant potential (also referred to as "high-risk" or B3 lesions). Finally, seven frequently asked questions are answered.

Abstract 593: Mammographic breast density in postmenopausal women is affected by the age of first full term pregnancy

Mammographic breast density (MBD) is accepted as a risk factor in breast cancer, whereas age of first full term pregnancy (FFTP) has been demonstrated to be a protective one. The objective of this study is to investigate the association between reproductive history and MBD at menopause. Women, age 50-69y, participating in the Flemish (Belgium) governmental organised breast cancer screening program were invited to participate in the study during their visit at the mammography department in Leuven (hospital and the mobile unit). The study was approved by the ethical committee of the hospital. After signing the informed consent, a self-administered questionnaire (SAQ) providing information on lifestyle and environmental exposure was completed by the participants. In the present work we present the analysis of the first 640 women that completed the SAQ and underwent mammography. Questionnaire data were registered via the Qualtrics software. The three parameters of MBD (VolparaDataManager®, USA) included the percentage glandular tissue of the total breast volume (GLAND), the ratio of the glandular tissue volume compared to the whole breast volume or Volumetric Breast Density (VBD) and the BIRADS density classification. We applied linear regression models to investigate the association between the continuous breast density markers (GLAND and VBD, log transformed) and the following explanatory variables: age of woman, ever pregnant, number of pregnancies, age of FFTP, ever breastfed (BF), amount and duration of BF, and breastfeeding in FFTP. An ordinal logistic regression model was used for the breast density marker BIRADS. The average GLAND and VBD changes with a multiplicative factor of 1.073 (p = 0.013) and 1.072 (p = 0.013) for a 5-year increase of age of FFTP. Thus, an increase of 5 years in age of FFTP is associated with about 7% increase in geometric mean of GLAND and VBD. The odds ratio for a higher versus a lower score of the BIRADS classification is 1.24 (p = 0.023) greater for every 5-year increase in age of FFTP. These results are independent for the inverse association between the age of the woman and the MBD markers. For a 5 year increment in age, geometric GLAND (p = 0.019) and VBD (p = 0.024) are approximately 5% lower and the odds ratio for a higher BIRADS score is 12% higher (p = 0.079). There is no association between the MBD markers and breastfeeding, the number and duration BF, and the number of pregnancies after correcting for age and age of FFTP. In conclusion, this study is to our knowledge the first one to show that MBD is significantly reduced when pregnancy takes place at early age.( This study was supported by a Grant of the Breast Cancer Organization Think Pink of Belgium, the Department of Breast Radiology and Mobile Unit of the University Hospital Leuven, the Center of Environmental Sciences and the Center of Biostatistics of the Hasselt University, and NIH Grant CA06927 to FCCC, PA, USA.)

Citation Format: Magda J. Vandeloo, Kristof Y. Neven, Liesbeth Bruckers, Jose Russo, Liesbeth Vancoullie, Esmée Bijnens, Eliane Kellen, Tim S. Nawrot, Chantal Van Ongeval. Mammographic breast density in postmenopausal women is affected by the age of first full term pregnancy [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 593.

Stromal characteristics are adequate prognosticators for recurrence risk in ductal carcinoma in situ of the breast

BACKGROUND

Ductal carcinoma in situ (DCIS) of the breast constitutes a heterogeneous group of non-obligate precursors for invasive breast cancer. To date, adequate risk stratification is lacking, which is presumed to result in overtreatment. We previously identified myxoid stromal architecture as a potential prognosticator for loco-regional recurrence. In the present study, we investigated the prognostic potential of stromal characteristics.

METHODS

Hematoxylin and eosin stained slides from 211 DCIS patients were reviewed. The following histological features were dichotomously assessed: nuclear grade, DCIS architecture, presence of necrosis, intraductal calcifications, stromal inflammation and myxoid stromal architecture. Loco-regional recurrences constituted the primary endpoint.

RESULTS

Cox regression analysis showed that high nuclear grade, myxoid stromal architecture and moderate to extensive stromal inflammation were significantly associated with decreased recurrence-free survival, independent of radiotherapy. Based on these features, a combined risk score (CRS) was calculated, ranging from zero to three. A high CRS of three was associated with significantly shorter recurrence-free survival. Nineteen patients had a CRS of three, of which three relapsed (15.7%), whereas only one out of 113 patients with a CRS of zero relapsed (0.9%).

CONCLUSIONS

We were able to validate our previously reported findings regarding the prognostic potential of myxoid periductal stroma in an independent DCIS patient cohort. A CRS based on nuclear grade, myxoid stromal architecture and stromal inflammation might facilitate discrimination of low risk from high risk patients. Consequently, the CRS may tailor adjuvant therapy. Future research should investigate whether radiotherapy can be safely omitted in patients with a low CRS.

Breast imaging surveillance after curative treatment for primary non-metastasised breast cancer in non-high-risk women: a systematic review

OBJECTIVES

The article summarises the available guidelines on breast imaging surveillance after curative treatment for locoregional breast cancer.

METHODS

A systematic review of practice guidelines published from 1 January 2007 to 1 January 2017 was performed according to PRISMA methodology. The search was conducted for the EMBASE, MEDLINE, Cochrane and Centre for Reviews and Dissemination databases. On 8 July 2018, all included guidelines were updated to the most recent version.

RESULTS

Twenty-one guidelines originating from 18 publishing bodies matched criteria. Publishing bodies consisted of seven governmental institutions, nine medical societies and two mixed collaborations. Publishing boards consisted of six radiological, four oncological, and 11 multidisciplinary teams. Annual bilateral mammography surveillance after breast-conserving therapy was recommended by 17/18 (94.4%) publishing bodies. Annual contralateral mammography surveillance after mastectomy was recommended by 13/18 (72.2%) publishing bodies. Routine use of digital breast tomosynthesis was recommended by 1/18 (5.6%) publishing bodies. Routine breast ultrasound surveillance was recommended by 2/18 (11.1%), deemed optional by 4/18 (22.2%) and not supported by 8/18 (44.4%) publishing bodies. Routine breast magnetic resonance imaging (MRI) surveillance was not recommended by 16/18 (88.9%) publishing bodies, although 6/18 (33.3%) specified subgroups for systematic MRI surveillance.

CONCLUSIONS

Annual mammography is currently the 'gold standard' for breast imaging surveillance. The role of digital breast tomosynthesis (DBT) remains to be further investigated. Most guidelines do not recommend routine breast ultrasound or MRI surveillance, unless indicated by additional risk factors.

Impact of compressed breast thickness and dose on lesion detectability in digital mammography: FROC study with simulated lesions in real mammograms

PURPOSE

The aim of this work was twofold: (1) to examine whether, with standard automatic exposure control (AEC) settings that maintain pixel values in the detector constant, lesion detectability in clinical images decreases as a function of breast thickness and (2) to verify whether a new AEC setup can increase lesion detectability at larger breast thicknesses.

METHODS

Screening patient images, acquired on two identical digital mammography systems, were collected over a period of 2 yr. Mammograms were acquired under standard AEC conditions (part 1) and subsequently with a new AEC setup (part 2), programmed to use the standard AEC settings for compressed breast thicknesses ≤49 mm, while a relative dose increase was applied above this thickness. The images were divided into four thickness groups: T1 ≤ 29 mm, T2 = 30-49 mm, T3 = 50-69 mm, and T4 ≥ 70 mm, with each thickness group containing 130 randomly selected craniocaudal lesion-free images. Two measures of density were obtained for every image: a BI-RADS score and a map of volumetric breast density created with a software application (VolparaDensity, Matakina, NZ). This information was used to select subsets of four images, containing one image from each thickness group, matched to a (global) BI-RADS score and containing a region with the same (local) volpara volumetric density value. One selected lesion (a microcalcification cluster or a mass) was simulated into each of the four images. This process was repeated so that, for a given thickness group, half the images contained a single lesion and half were lesion-free. The lesion templates created and inserted in groups T3 and T4 for the first part of the study were then inserted into the images of thickness groups T3 and T4 acquired with higher dose settings. Finally, all images were visualized using the ViewDEX software and scored by four radiologists performing a free search study. A statistical jackknife-alternative free-response receiver operating characteristic analysis was applied.

RESULTS

For part 1, the alternative free-response receiver operating characteristic curves for the four readers were 0.80, 0.65, 0.55 and 0.56 in going from T1 to T4, indicating a decrease in detectability with increasing breast thickness. P-values and the 95% confidence interval showed no significant difference for the T3-T4 comparison (p = 0.78) while all the other differences were significant (p < 0.05). Separate analysis of microcalcification clusters presented the same results while for mass detection, the only significant difference came when comparing T1 to the other thickness groups. Comparing the scores of part 1 and part 2, results for the T3 group acquired with the new AEC setup and T3 group at standard AEC doses were significantly different (p = 0.0004), indicating improved detection. For this group a subanalysis for microcalcification detection gave the same results while no significant difference was found for mass detection.

CONCLUSIONS

These data using clinical images confirm results found in simple QA tests for many mammography systems that detectability falls as breast thickness increases. Results obtained with the AEC setup for constant detectability above 49 mm showed an increase in lesion detection with compressed breast thickness, bringing detectability of lesions to the same level.

Screen-detected versus interval cancers: Effect of imaging modality and breast density in the Flemish Breast Cancer Screening Programme

OBJECTIVES

To investigate if direct radiography (DR) performs better than screen-film mammography (SF) and computed radiography (CR) in dense breasts in a decentralized organised Breast Cancer Screening Programme. To this end, screen-detected versus interval cancers were studied in different BI-RADS density classes for these imaging modalities.

METHODS

The study cohort consisted of 351,532 women who participated in the Flemish Breast Cancer Screening Programme in 2009 and 2010. Information on screen-detected and interval cancers, breast density scores of radiologist second readers, and imaging modality was obtained by linkage of the databases of the Centre of Cancer Detection and the Belgian Cancer Registry.

RESULTS

Overall, 67% of occurring breast cancers are screen detected and 33% are interval cancers, with DR performing better than SF and CR. The interval cancer rate increases gradually with breast density, regardless of modality. In the high-density class, the interval cancer rate exceeds the cancer detection rate for SF and CR, but not for DR.

CONCLUSIONS

DR is superior to SF and CR with respect to cancer detection rates for high-density breasts. To reduce the high interval cancer rate in dense breasts, use of an additional imaging technique in screening can be taken into consideration.

KEY POINTS

• Interval cancer rate increases gradually with breast density, regardless of modality. • Cancer detection rate in high-density breasts is superior in DR. • IC rate exceeds CDR for SF and CR in high-density breasts. • DR performs better in high-density breasts for third readings and false-positives.

Prognostic features of breast cancer differ between women in the Democratic Republic of Congo and Belgium

OBJECTIVES

Compared to European women, breast cancers in African women present at a younger age, with a higher tumor grade and are more often estrogen receptor (ER)/progesterone receptor (PR) negative. We here investigate the histopathological and immunohistochemical characteristics (ER, PR and human epidermal growth receptor 2 (HER2)) and the proportion of triple negative (Tneg) invasive breast cancers from an unselected series of patients diagnosed in Kinshasa, and compare them to a population of Caucasian women with a palpable breast cancer.

MATERIALS AND METHODS

From 2010 till 2013, during the first breast cancer awareness campaign, organized in Kinshasa, 87 patients were diagnosed with invasive breast cancer. Diagnose was based on core biopsy. The control group consisted of Caucasian women (University Hospitals of Leuven, Belgium) with a palpable mass, diagnosed between 2000 till 2009, treated with surgery of which the histopathological and immunohistochemical characteristics were collected on excision specimens. Each patient in the Kinshasa group was matched based on age and tumor size to one or more patients of the Leuven database. Differences between both groups with respect to hormone receptors (ER, PR, HER2, Tneg) or grade are presented as relative risks (RR). The analysis is based on a log-binomial model accounting for clustering through matching by a random intercept for cluster. Differences between both groups with respect to hormone receptors correcting for grade is performed by the inclusion of grade as a covariate in the model.

RESULTS

After adjusting for age, tumor volume and tumor grade, ER was more frequently negative (RR = 0.71, p < 0.001), with a trend in the same direction for PR (RR = 0.87, p = 0.057), and HER2 more often positive (RR = 1.60, p = 0.015) compared to the group from the University Hospitals of Leuven. There was no difference in the proportion of breast cancers being triple negative. Sub-analysis showed that the higher HER2 positive rate was only observed in older patients (≥50y: RR = 2.07, p = 0.007) whereas no difference in HER2 positive rate was found in younger patients (<50y: RR = 1.30, p = 0.358). A higher ER negative rate was observed in both age groups, however more pronounced in older patients (≥50y: RR = 0.64, p = 0.001; <50y: RR = 0.79, p = 0.018).

CONCLUSION

Breast cancer in women of Kinshasa presents at younger age and is more aggressive (more frequently ER negative and HER2 positive) compared to Caucasian women and this is more pronounced in older women (>50y). Only the ER results were concordant with the results of two similar studies (comparing an African with a European group), but were different when compared to studies on African-American women with breast cancer. This information is very important considering the treatment option: as more tumors are ER negative, endocrine therapy cannot be given while chemotherapy is often too expensive.

Comparison of digital breast tomosynthesis and 2D digital mammography using a hybrid performance test

This paper introduces a hybrid method for performing detection studies in projection image based modalities, based on image acquisitions of target objects and patients. The method was used to compare 2D mammography and digital breast tomosynthesis (DBT) in terms of the detection performance of spherical densities and microcalcifications. The method starts with the acquisition of spheres of different glandular equivalent densities and microcalcifications of different sizes immersed in a homogeneous breast tissue simulating medium. These target objects are then segmented and the subsequent templates are fused in projection images of patients and processed or reconstructed. This results in hybrid images with true mammographic anatomy and clinically relevant target objects, ready for use in observer studies. The detection study of spherical densities used 108 normal and 178 hybrid 2D and DBT images; 156 normal and 321 hybrid images were used for the microcalcifications. Seven observers scored the presence/absence of the spheres/microcalcifications in a square region via a 5-point confidence rating scale. Detection performance in 2D and DBT was compared via ROC analysis with sub-analyses for the density of the spheres, microcalcification size, breast thickness and z-position. The study was performed on a Siemens Inspiration tomosynthesis system using patient acquisitions with an average age of 58 years and an average breast thickness of 53 mm providing mean glandular doses of 1.06 mGy (2D) and 2.39 mGy (DBT). Study results showed that breast tomosynthesis (AUC = 0.973) outperformed 2D (AUC = 0.831) for the detection of spheres (p  <  0.0001) and this applied for all spherical densities and breast thicknesses. By way of contrast, DBT was worse than 2D for microcalcification detection (AUC2D = 0.974, AUCDBT = 0.838, p  <  0.0001), with significant differences found for all sizes (150-354 µm), for breast thicknesses above 40 mm and for heights above the detector of 20 mm and above. In conclusion, the hybrid method was successfully used to produce images for a detection study; results showed breast tomosynthesis outperformed 2D for spherical densities while further optimization of DBT for microcalcifications is suggested.

Prediction of breast cancer risk based on profiling with common genetic variants

BACKGROUND

Data for multiple common susceptibility alleles for breast cancer may be combined to identify women at different levels of breast cancer risk. Such stratification could guide preventive and screening strategies. However, empirical evidence for genetic risk stratification is lacking.

METHODS

We investigated the value of using 77 breast cancer-associated single nucleotide polymorphisms (SNPs) for risk stratification, in a study of 33 673 breast cancer cases and 33 381 control women of European origin. We tested all possible pair-wise multiplicative interactions and constructed a 77-SNP polygenic risk score (PRS) for breast cancer overall and by estrogen receptor (ER) status. Absolute risks of breast cancer by PRS were derived from relative risk estimates and UK incidence and mortality rates.

RESULTS

There was no strong evidence for departure from a multiplicative model for any SNP pair. Women in the highest 1% of the PRS had a three-fold increased risk of developing breast cancer compared with women in the middle quintile (odds ratio [OR] = 3.36, 95% confidence interval [CI] = 2.95 to 3.83). The ORs for ER-positive and ER-negative disease were 3.73 (95% CI = 3.24 to 4.30) and 2.80 (95% CI = 2.26 to 3.46), respectively. Lifetime risk of breast cancer for women in the lowest and highest quintiles of the PRS were 5.2% and 16.6% for a woman without family history, and 8.6% and 24.4% for a woman with a first-degree family history of breast cancer.

CONCLUSIONS

The PRS stratifies breast cancer risk in women both with and without a family history of breast cancer. The observed level of risk discrimination could inform targeted screening and prevention strategies. Further discrimination may be achievable through combining the PRS with lifestyle/environmental factors, although these were not considered in this report.

The simulation of 3D mass models in 2D digital mammography and breast tomosynthesis

PURPOSE

This work proposes a new method of building 3D breast mass models with different morphological shapes and describes the validation of the realism of their appearance after simulation into 2D digital mammograms and breast tomosynthesis images.

METHODS

Twenty-five contrast enhanced MRI breast lesions were collected and each mass was manually segmented in the three orthogonal views: sagittal, coronal, and transversal. The segmented models were combined, resampled to have isotropic voxel sizes, triangularly meshed, and scaled to different sizes. These masses were referred to as nonspiculated masses and were then used as nuclei onto which spicules were grown with an iterative branching algorithm forming a total of 30 spiculated masses. These 55 mass models were projected into 2D projection images to obtain mammograms after image processing and into tomographic sequences of projection images, which were then reconstructed to form 3D tomosynthesis datasets. The realism of the appearance of these mass models was assessed by five radiologists via receiver operating characteristic (ROC) analysis when compared to 54 real masses. All lesions were also given a breast imaging reporting and data system (BIRADS) score. The data sets of 2D mammography and tomosynthesis were read separately. The Kendall's coefficient of concordance was used for the interrater observer agreement assessment for the BIRADS scores per modality. Further paired analysis, using the Wilcoxon signed rank test, of the BIRADS assessment between 2D and tomosynthesis was separately performed for the real masses and for the simulated masses.

RESULTS

The area under the ROC curves, averaged over all observers, was 0.54 (95% confidence interval [0.50, 0.66]) for the 2D study, and 0.67 (95% confidence interval [0.55, 0.79]) for the tomosynthesis study. According to the BIRADS scores, the nonspiculated and the spiculated masses varied in their degrees of malignancy from normal (BIRADS 1) to highly suggestive for malignancy (BIRADS 5) indicating the required variety of shapes and margins of these models. The assessment of the BIRADS scores for all observers indicated good agreement based on Kendall's coefficient for both the 2D and the tomosynthesis evaluations. The paired analysis of the BIRADS scores between 2D and tomosynthesis for each observer revealed consistent behavior for the real and simulated masses.

CONCLUSIONS

A database of 3D mass models, with variety of shapes and margins, was validated for the realism of their appearance for 2D digital mammography and for breast tomosynthesis. This database is suitable for use in future observer performance studies whether in virtual clinical trials or in patient images with simulated lesions.

From the set-up of a screening program of breast cancer patients to the identification of the first BRCA mutation in the DR Congo

BACKGROUND

Breast cancer incidence in African population is low compared to western countries but the mortality rate is higher and the disease presents at a younger age and at a more advanced stage. The World Health Organisation and the Breast Health Global Initiative concluded that in low and middle income countries early breast cancer detection can be achieved by informing women on symptoms of breast cancer, on the practice of breast self-examination and clinical breast examination by trained health care workers. Based on these recommendations, we set up a breast cancer awareness campaign in Kinshasa, Democratic Republic of Congo (DRC). This paper describes the strategy that was established and the results that were achieved.

METHODS

A breast cancer awareness campaign was started in 2010 and data were collected until the end of 2012. Clinicians (expert group) trained nurses and health care workers (awareness groups) on clinical, technical and social aspects of breast cancer. Different channels were used to inform women about the campaign and clinical data (on medical and family history) were collected. The participating women were investigated with clinical breast examination by the awareness group. Women in whom a palpable mass was detected were referred to the hospital: they received a mammography and ultrasound and--in case of suspicious findings--additionally a core needle biopsy. In case of a positive family history, a blood sample was taken for genetic investigation.

RESULTS

In total, 4,315 women participated, resulting in 1,113 radiological breast examinations, performed in the General Hospital of Kinshasa of which 101 turned out to be malignant lesions. Fifty six percent of the women with breast cancer were less than 50 years old and 75% (65/87) were stage III tumors. A BRCA gene mutation was identified in a family with a severe history of breast cancer.

CONCLUSIONS

Even without financial support, it was possible to start an awareness campaign for breast cancer in Kinshasa. This campaign increased the awareness on cancer of the women in Kinshasa. The results demonstrate that this campaign had an immediate impact on patients and their families.

From the set-up of a screening program of breast cancer patients to the identification of the first BRCA mutation in the DR Congo

BACKGROUND

Breast cancer incidence in African population is low compared to western countries but the mortality rate is higher and the disease presents at a younger age and at a more advanced stage. The World Health Organisation and the Breast Health Global Initiative concluded that in low and middle income countries early breast cancer detection can be achieved by informing women on symptoms of breast cancer, on the practice of breast self-examination and clinical breast examination by trained health care workers. Based on these recommendations, we set up a breast cancer awareness campaign in Kinshasa, Democratic Republic of Congo (DRC). This paper describes the strategy that was established and the results that were achieved.

METHODS

A breast cancer awareness campaign was started in 2010 and data were collected until the end of 2012. Clinicians (expert group) trained nurses and health care workers (awareness groups) on clinical, technical and social aspects of breast cancer. Different channels were used to inform women about the campaign and clinical data (on medical and family history) were collected. The participating women were investigated with clinical breast examination by the awareness group. Women in whom a palpable mass was detected were referred to the hospital: they received a mammography and ultrasound and--in case of suspicious findings--additionally a core needle biopsy. In case of a positive family history, a blood sample was taken for genetic investigation.

RESULTS

In total, 4,315 women participated, resulting in 1,113 radiological breast examinations, performed in the General Hospital of Kinshasa of which 101 turned out to be malignant lesions. Fifty six percent of the women with breast cancer were less than 50 years old and 75% (65/87) were stage III tumors. A BRCA gene mutation was identified in a family with a severe history of breast cancer.

CONCLUSIONS

Even without financial support, it was possible to start an awareness campaign for breast cancer in Kinshasa. This campaign increased the awareness on cancer of the women in Kinshasa. The results demonstrate that this campaign had an immediate impact on patients and their families.

Is the sentinel lymph node biopsy more sensitive for the identification of positive lymph nodes in breast cancer than the axillary lymph node dissection?

Since the routine clinical use of the sentinel lymph node (SLN) procedure, questions have been raised concerning an increase in the overall percentage of node-positive patients. The goal of our study was to compare the sensitivity of the SLN procedure and the axillary lymph node dissection (ALND) for the identification of positive lymph nodes in breast cancer.

The incidence of axillary node metastasis in SLNB and ALND specimens from patients undergoing operative treatment of a primary breast carcinoma was compared retrospectively.

Logistic regression models were used to analyze the effect of various predictors on the presence of positive lymph nodes. We constructed a multivariate model including the procedure and these predictors that have shown to be related to lymph node involvement in univariate analysis. The probability of finding positive lymph nodes was thus calculated in both groups correcting for relevant predictors of lymph node involvement.

The SLNB group included 830 patients, the ALND group 320. In a multivariate analysis, adjusting for the number of foci, tumor location in the breast, tumor size, LVI, ER, PR, tumor grade and histological subtype, the probability of finding positive lymph nodes was higher with SLNB procedure than with an ALND. However, this difference was not statistically significant (OR 0.7635; CI 0.5334-1.0930, p 0.1404).

For comparable tumors, SLNB procedure is at least as sensitive as ALND for detecting positive lymph nodes.

Technical and clinical breast cancer screening performance indicators for computed radiography versus direct digital radiography

OBJECTIVES

To compare technical and clinical screening performance parameters between computed radiography (CR) and direct digital radiography (DR) systems.

METHODS

The number of women screened with CR was 73,008 and with DR 116,945. Technical and patient dose survey data of 25 CR and 37 DR systems were available. Technical performance was expressed by threshold thickness values at the mean glandular dose (MGD) level of routine practice. Clinical indicators included recall rate (RR), cancer detection rate (CDR), percentage of ductal carcinoma in situ (DCIS), percentage of cancers with T-scores smaller than 1 cm and positive predictive value (PPV).

RESULTS

Contrast threshold values for the 0.1-mm gold disk were 1.44 μm (SD 0.13 μm) for CR and 1.20 μm (SD 0.13 μm for DR). MGD was 2.16 mGy (SD 0.36 mGy) and 1.35 mGy (SD 0.32 mGy) for CR and DR respectively. We obtained for CR, respectively DR, the following results: RR in the first round of 5.48 % versus 5.61 %; RR in subsequent rounds of 2.52 % versus 2.65 %; CDR of 0.52 % versus 0.53 %; DCIS of 0.08 % versus 0.11 %; a rate of cancers with T-scores smaller than 1 cm of 0.11 % versus 0.11 %; PPV of 18.45 % versus 18.64 %; none of them was significantly different.

CONCLUSION

Our screening indicators are reassuring for the use of CR and DR, with CR operating at 60 % higher MGD.

KEY POINTS

• Breast cancer screening can employ both computed (CR) and direct digital radiography (DR). • Screening performance parameters for CR and DR technology are not significantly different. • Screening parameters are in accordance with European Guidelines. • Radiation doses employed for CR are generally 60 % greater than for DR.

Correlation of free-response and receiver-operating-characteristic area-under-the-curve estimates: results from independently conducted FROC∕ROC studies in mammography

PURPOSE

From independently conducted free-response receiver operating characteristic (FROC) and receiver operating characteristic (ROC) experiments, to study fixed-reader associations between three estimators: the area under the alternative FROC (AFROC) curve computed from FROC data, the area under the ROC curve computed from FROC highest rating data, and the area under the ROC curve computed from confidence-of-disease ratings.

METHODS

Two hundred mammograms, 100 of which were abnormal, were processed by two image-processing algorithms and interpreted by four radiologists under the FROC paradigm. From the FROC data, inferred-ROC data were derived, using the highest rating assumption. Eighteen months afterwards, the images were interpreted by the same radiologists under the conventional ROC paradigm; conventional-ROC data (in contrast to inferred-ROC data) were obtained. FROC and ROC (inferred, conventional) data were analyzed using the nonparametric area-under-the-curve (AUC), (AFROC and ROC curve, respectively). Pearson correlation was used to quantify the degree of association between the modality-specific AUC indices and standard errors were computed using the bootstrap-after-bootstrap method. The magnitude of the correlations was assessed by comparison with computed Obuchowski-Rockette fixed reader correlations.

RESULTS

Average Pearson correlations (with 95% confidence intervals in square brackets) were: Corr(FROC, inferred ROC) = 0.76[0.64, 0.84] > Corr(inferred ROC, conventional ROC) = 0.40[0.18, 0.58] > Corr (FROC, conventional ROC) = 0.32[0.16, 0.46].

CONCLUSIONS

Correlation between FROC and inferred-ROC data AUC estimates was high. Correlation between inferred- and conventional-ROC AUC was similar to the correlation between two modalities for a single reader using one estimation method, suggesting that the highest rating assumption might be questionable.

Predictors of axillary lymph node metastases in early breast cancer and their applicability in clinical practice

PURPOSE

Lymph node involvement is the most important prognostic factor in breast cancer. It is a multifactorial event determined by patient and tumour characteristics. The purpose of this study was to determine clinical and pathological factors predictive for axillary lymph node metastasis (ALNM) in patients with early breast cancer and to build a model to portend lymph node involvement.

METHODS

We evaluated 1300 consecutive patients surgically treated in our institution (2007-2009) for cT1-T2 invasive breast cancer. The patient and tumour characteristics evaluated included: age at diagnosis, number of foci, histologic grade, location, tumour size, histologic subtype, lymphovascular invasion (LVI), estrogen-receptor (ER), progesterone-receptor (PR) and Her-2 status. Univariate and multivariate analyses were performed. Factors significantly associated with ALNM by univariate analysis plus histologic subtype were included in the multivariate analysis.

RESULTS

By univariate analysis, the incidence of ALNM was significantly associated with the presence of LVI (P < 0.0001), larger tumour size (P < 0.0001), higher histologic grade (P < 0.0001), retroareolar or lateral location in the breast (P < 0.0001), multiple foci (P = 0.0002) and in patients who underwent an axillary lymph node dissection. We found no effect of age, ER⁄PR nor HER-2 status. By multivariate analysis, ALNM was significantly associated with the presence of LVI (P < 0.0001), larger tumour size (P < 0.0001), axillary lymph node dissection (P = 0.0003), retroareolar and lateral tumour location in the breast (P = 0.0019) and the presence of multiple foci (P = 0.0155).

CONCLUSIONS

LVI and tumour size emerged as the most powerful independent predictors of ALNM, followed by the location of the tumour in the breast and the presence of multiple foci.

The simulation of 3D microcalcification clusters in 2D digital mammography and breast tomosynthesis

PURPOSE

This work proposes a new method of building 3D models of microcalcification clusters and describes the validation of their realistic appearance when simulated into 2D digital mammograms and into breast tomosynthesis images.

METHODS

A micro-CT unit was used to scan 23 breast biopsy specimens of microcalcification clusters with malignant and benign characteristics and their 3D reconstructed datasets were segmented to obtain 3D models of microcalcification clusters. These models were then adjusted for the x-ray spectrum used and for the system resolution and simulated into 2D projection images to obtain mammograms after image processing and into tomographic sequences of projection images, which were then reconstructed to form 3D tomosynthesis datasets. Six radiologists were asked to distinguish between 40 real and 40 simulated clusters of microcalcifications in two separate studies on 2D mammography and tomosynthesis datasets. Receiver operating characteristic (ROC) analysis was used to test the ability of each observer to distinguish between simulated and real microcalcification clusters. The kappa statistic was applied to assess how often the individual simulated and real microcalcification clusters had received similar scores ("agreement") on their realistic appearance in both modalities. This analysis was performed for all readers and for the real and the simulated group of microcalcification clusters separately. "Poor" agreement would reflect radiologists' confusion between simulated and real clusters, i.e., lesions not systematically evaluated in both modalities as either simulated or real, and would therefore be interpreted as a success of the present models.

RESULTS

The area under the ROC curve, averaged over the observers, was 0.55 (95% confidence interval [0.44, 0.66]) for the 2D study, and 0.46 (95% confidence interval [0.29, 0.64]) for the tomosynthesis study, indicating no statistically significant difference between real and simulated lesions (p > 0.05). Agreement between allocated lesion scores for 2D mammography and those for the tomosynthesis series was poor.

CONCLUSIONS

The realistic appearance of the 3D models of microcalcification clusters, whether malignant or benign clusters, was confirmed for 2D digital mammography images and the breast tomosynthesis datasets; this database of clusters is suitable for use in future observer performance studies related to the detectability of microcalcification clusters. Such studies include comparing 2D digital mammography to breast tomosynthesis and comparing different reconstruction algorithms.

Evaluation of clinical image processing algorithms used in digital mammography

Screening is the only proven approach to reduce the mortality of breast cancer, but significant numbers of breast cancers remain undetected even when all quality assurance guidelines are implemented. With the increasing adoption of digital mammography systems, image processing may be a key factor in the imaging chain. Although to our knowledge statistically significant effects of manufacturer-recommended image processings have not been previously demonstrated, the subjective experience of our radiologists, that the apparent image quality can vary considerably between different algorithms, motivated this study. This article addresses the impact of five such algorithms on the detection of clusters of microcalcifications. A database of unprocessed (raw) images of 200 normal digital mammograms, acquired with the Siemens Novation DR, was collected retrospectively. Realistic simulated microcalcification clusters were inserted in half of the unprocessed images. All unprocessed images were subsequently processed with five manufacturer-recommended image processing algorithms (Agfa Musica 1, IMS Raffaello Mammo 1.2, Sectra Mamea AB Sigmoid, Siemens OPVIEW v2, and Siemens OPVIEW v1). Four breast imaging radiologists were asked to locate and score the clusters in each image on a five point rating scale. The free-response data were analyzed by the jackknife free-response receiver operating characteristic (JAFROC) method and, for comparison, also with the receiver operating characteristic (ROC) method. JAFROC analysis revealed highly significant differences between the image processings (F = 8.51, p < 0.0001), suggesting that image processing strongly impacts the detectability of clusters. Siemens OPVIEW2 and Siemens OPVIEW1 yielded the highest and lowest performances, respectively. ROC analysis of the data also revealed significant differences between the processing but at lower significance (F = 3.47, p = 0.0305) than JAFROC. Both statistical analysis methods revealed that the same six pairs of modalities were significantly different, but the JAFROC confidence intervals were about 32% smaller than ROC confidence intervals. This study shows that image processing has a significant impact on the detection of microcalcifications in digital mammograms. Objective measurements, such as described here, should be used by the manufacturers to select the optimal image processing algorithm.

An improved method for simulating microcalcifications in digital mammograms

The assessment of the performance of a digital mammography system requires an observer study with a relatively large number of cases with known truth which is often difficult to assemble. Several investigators have developed methods for generating hybrid abnormal images containing simulated microcalcifications. This article addresses some of the limitations of earlier methods. The new method is based on digital images of needle biopsy specimens. Since the specimens are imaged separately from the breast, the microcalcification attenuation profile scan is deduced without the effects of over and underlying tissues. The resulting templates are normalized for image acquisition specific parameters and reprocessed to simulate microcalcifications appropriate to other imaging systems, with different x-ray, detector and image processing parameters than the original acquisition system. This capability is not shared by previous simulation methods that have relied on extracting microcalcifications from breast images. The method was validated by five experienced mammographers who compared 59 pairs of simulated and real microcalcifications in a two-alternative forced choice task designed to test if they could distinguish the real from the simulated lesions. They also classified the shapes of the microcalcifications according to a standardized clinical lexicon. The observed probability of correct choice was 0.415, 95% confidence interval (0.284, 0.546), showing that the radiologists were unable to distinguish the lesions. The shape classification revealed substantial agreement with the truth (mean kappa = 0.70), showing that we were able to accurately simulate the lesion morphology. While currently limited to single microcalcifications, the method is extensible to more complex clusters of microcalcifications and to three-dimensional images. It can be used to objectively assess an imaging technology, especially with respect to its ability to adequately visualize the morphology of the lesions, which is a critical factor in the benign versus malignant classification of a lesion detected in screening mammography.

Clinical image quality criteria for full field digital mammography: a first practical application

In order to quantify the clinical quality of full-field digital mammography, a set of image quality parameters is developed. The set consisted of 12 image quality criteria and 8 physical characteristics of the image. The first set interrogates the visibility of anatomical structures and typical characteristics of a digital image, such as noise and saturation of dark and white areas. The second set of criteria evaluates contrast, sharpness and confidence with the representation of masses, microcalcifications and the image. The use of these criteria is reported in a retrospective study, in which the impact of dose on the radiological quality of digital mammograms is evaluated. Fifty patients acquired in a low-dose mode were retrieved and compared with 50 patients acquired in a dose mode that was set 41% higher. The dose affects, more than expected, contrast and sharpness of the image, whereas the visibility of the anatomical structures remains unchanged. With these parameters, quantification of the image quality is possible; however, because of subjectivity of the parameters, only intra-observer comparison and evaluation of the individual parameters rather than the overall results are advised. Together with physical tests of image quality, critical radiological evaluation of the quality should be included in the acceptance process of digital mammography.

Teaching syllabus for radiological aspects of breast cancer screening with digital mammography

The purpose of this study is to discuss the content of our new accreditation programme for radiologists' reading digital mammograms in a screening setting and to report our first experience with the new course. The course consisted of a theoretical part, given by the medical physicist, and a practical part given by the radiologist. The practical session is closely linked with the theoretical part and a reading session. The material is fully digital and can be presented on different platforms. In practice, the need for parallel soft-copy reading sessions on high-end workstations limits the number of participants. A high level of interactivity was noted between teacher and participant, with a thorough discussion of different digital mammography systems during a single teaching course. The main challenge for the teacher turned out to be the collection of representative material and the continuous updating of the material: new systems, processing techniques and artefacts need to be included regularly.

Prophylactic mastectomy in familial breast carcinoma. What do the pathologic findings learn us?

OBJECTIVE

To evaluate the indications, techniques and pathologic findings of prophylactic mastectomy (pME).

METHODS

Retrospective study of patients with a strong family history of breast cancer (and ovarian cancer) or with proven BRCA mutation.

RESULTS

Between January 1995 and December 2005, sixty seven patients underwent pME: 15% had a BRCA1 mutation, 31% had a BRCA2 mutation and 33% had a strong family history but without proven BRCA mutation and in 21% mutation analysis was not performed. Fifty eight percent had a personal history of breast cancer of which 84% previously underwent a unilateral mastectomy as part of their treatment. The median time to decision from previous treatment for breast carcinoma to pME was 46 months. Mean age at pME was 43 years. Pathologic examination of the pME specimens revealed invasive and/or in situ carcinoma in 19% (13/67). Atypical ductal/lobular hyperplasia (ADH/ALH) or flat epithelial atypia (FEA) were present in another 3%.

CONCLUSION

Twenty two percent of women at high risk for breast cancer presented at the time of pME with invasive carcinoma or intra-epithelial neoplasia undetected by imaging and clinical examination.

Current status of digital mammography for screening and diagnosis of breast cancer

PURPOSE OF REVIEW

Digital mammography is becoming the preferred technique for investigation of the breast. It is important, therefore, to analyze not only the accuracy of this technique in the screening and the diagnostic environment but also to evaluate its strengths and limits. In addition, communication with other specialists inside and outside the hospital is essential.

RECENT FINDINGS

Recent publications of large clinical trials have shown that digital mammography is as accurate as film-screen mammography in a screening setting. Technical protocols for acceptance testing of these modalities are emerging but they are not yet complete. The literature shows that reading on soft copy may be preferred to hard copy. The high cost is still an important limiting factor in the easy introduction of full-field digital mammography in a hospital.

SUMMARY

Digital mammography is becoming the method of choice in the detection and characterization of breast cancer. Today, physical and technical protocols as well as large clinical trials are assessing the performance of this technology. A lot of work remains in the optimization of the different parts of the imaging chain: exposure setting, acceptability of detectors, dedicated post processing, viewing conditions and computer-aided detection. In parallel with these developments, newer digital technologies are being explored (tomosynthesis).

Quantification of Al-equivalent thickness of just visible microcalcifications in full field digital mammograms

Characterization of digital mammography systems is often performed by means of contrast-detail curves using a homogeneous phantom with inserts of different sizes and thicknesses. In this article, a more direct measure of the threshold contrast-detail characteristics of microcalcifications in clinical mammograms is proposed, which also takes into account routine processing and display. The proposed method scores the detectability of simulated microcalcifications with known size and aluminum-equivalent thickness. Thickness estimates, based on x-ray transmission coefficients, were first validated for Al particles. The same approach was then applied to associate Al-equivalent thickness with simulated microcalcifications. Thirty-five mammograms of patients were acquired using a full field digital mammography (FFDM) system operating under standard exposure conditions. Different microcalcifications were simulated using templates of real microcalcifications as described in Med. Phys. 30, 2234-2240 (2003). These templates were first modified such that they simulated a template of the same microcalcification for an ideally sharp detector. They were then adjusted for the imaging characteristics of the FFDM, beam quality, and breast thickness. Microcalcification sizes in the image plane ranged from 200 to 800 microm. Their peak Al-equivalent thickness varied between 70 and 1000 microm. Software phantoms were created. They consisted of 0-10 simulated microcalcifications randomly distributed in 2 cm by 2 cm frames embedded within digital mammograms. Routine processing and printing followed. Three experienced radiologists recorded the locations of the microcalcifications, and confidence ratings were given. Free response receiver operating characteristics (FROC) analysis was performed. Using a binary score, the fractions of detected microcalcifications were plotted as a function of equivalent diameter for the different Al-equivalent thicknesses. Pair-wise agreement of the detected microcalcifications was calculated for the different Al-equivalent thickness groups. The FROC curves of each radiologist indicated similar true positive fractions for a given number of false positives per image. One radiologist applied a more conservative scoring. Detected fractions for the different sizes of the microcalcifications showed the same trend for all observers. In addition, the observer with the least FP also detected less microcalcifications. The pair-wise agreement of the detected microcalcifications was good. The average detected fractions were >0.5 for microcalcifications with equivalent diameter >400 microm and Al-equivalent thickness >400 microm. An average detected fraction >0.5 was also seen for microcalcifications with equivalent diameter <400 microm and equivalent thickness >800 microm. The detected fractions of smaller microcalcifications were <0.5. The results obtained with this method indicate that it may be possible to quantify the performance of a digital mammography detector including processing and viewing for the detection of microcalcifications. We hypothesize that the FROC curves and detected fractions of simulated microcalcifications of different sizes reflect the clinical reality.

Development and validation of a simulation procedure to study the visibility of micro calcifications in digital mammograms

The visibility of micro calcifications is a determining factor for digital mammography. To address the problem of quantification, we developed a procedure to simulate micro calcifications into real mammograms. First, the shapes, sizes and x-ray transmission coefficients of real micro calcifications were derived from the appearance of biopsy specimens in the raw data of magnified, digital images acquired at 27 kVp and Mo/Mo anode-filter combination. This allowed us to create "ideal templates" of micro calcifications. The x-ray transmissions of the real micro calcifications values were expressed in Al-equivalent thickness. This made it possible to recalculate the x-ray transmission characteristics of a particular ideal template for other x-ray beam qualities. Extra corrections for differences in spatial resolution were based on the presampled two-dimensional modulation transfer functions and on the difference in pixel size. Three radiologists compared the appearance of real and simulated micro calcifications in a two-alternative forced choice (2AFC) evaluation. They perceived no differences between real and simulated lesions. Preliminary results show that it is possible to simulate micro calcifications with well defined characteristics that are indistinguishable from real ones. It should be noted, however, that the full potential of the approach has not been proven. In future work, these templates may be useful to evaluate particular aspects of digital mammograms, such as the effects of processing and of viewing conditions on the visibility of micro calcifications.